JMIR Publications

ABSTRACT

Background:

The assessment of Osteonecrosis of the Femoral Head (ONFH) often presents challenges in accuracy and efficiency. Traditional methods rely on imaging studies and clinical judgment, prompting the need for advanced approaches. This study aims to utilize deep learning algorithms to enhance disease assessment and prediction in ONFH, optimizing treatment strategies.

Objective:

The primary objective of this research is to analyze pathological images of ONFH using advanced deep learning algorithms to evaluate treatment response, vascular reconstruction, and disease progression. By identifying the most effective algorithm, this study seeks to equip clinicians with precise tools for disease assessment and prediction.

Methods:

MRI images from ONFH patients were collected, standardized, and annotated, forming datasets for model training and validation. Ten deep learning algorithms were tested and optimized, with MobileNetV3_Large selected as the best-performing model. This optimal model was used to extract treatment-related features, quantify treatment effects, and conduct statistical analyses. Additionally, vascular features were automatically segmented from the images to assess the impact of treatments on vascular reconstruction. A time-series dataset was created using pathological images from different time points to train a predictive model and ensure accuracy.

Results:

MobileNetV3_Large emerged as the optimal model for evaluating treatment response, vascular reconstruction, and disease progression in ONFH. It showcased superior performance in accuracy, recall, and F1 score compared to other algorithms, particularly in tracking the progression dynamics of ONFH. The model effectively quantified the impact of different treatments on affected regions, providing crucial decision-making support for clinicians. The integration of the MobileNetV3_Large model with an LSTM network facilitated accurate forecasting of disease progression, exhibiting a high level of predictive performance and robustness with an AUC of 0.92.

Conclusions:

The application of deep learning algorithms in examining treatment response, vascular reconstruction, and disease progression in ONFH presents notable advantages. This study offers clinicians a precise tool for disease assessment and highlights the significance of utilizing advanced technological solutions in healthcare practice.